U.S. patent number 7,777,476 [Application Number 12/139,551] was granted by the patent office on 2010-08-17 for dynamic modulation for multiplexation of microfluidic and nanofluidic based biosensors.
This patent grant is currently assigned to The University of Akron. Invention is credited to Jun Hu, Jiang Zhe.
United States Patent |
7,777,476 |
Hu , et al. |
August 17, 2010 |
Dynamic modulation for multiplexation of microfluidic and
nanofluidic based biosensors
Abstract
The present invention generally relates to a method for rapidly
counting micron and/or submicron particles by passing such
particles through any of a plurality of microfluidic channels
simultaneously with an ion current and measuring the signal
generated thereby. The present invention also generally relates to
a device for practicing the method of the present invention. Some
embodiments can include methods and/or devices for distinguishing
between and counting particles in mixtures. Still other embodiments
can include methods and/or devices for identifying and/or counting
bioparticles and/or bioactive particles such as pollen.
Inventors: |
Hu; Jun (Fairlawn, OH), Zhe;
Jiang (Copley, OH) |
Assignee: |
The University of Akron (Akron,
OH)
|
Family
ID: |
40431171 |
Appl.
No.: |
12/139,551 |
Filed: |
June 16, 2008 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20090066315 A1 |
Mar 12, 2009 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
11584945 |
Oct 23, 2006 |
7397232 |
|
|
|
60729262 |
Oct 21, 2005 |
|
|
|
|
Current U.S.
Class: |
324/71.4;
324/71.1; 73/865.5 |
Current CPC
Class: |
G01N
15/12 (20130101); G01N 15/1056 (20130101); G01N
15/1245 (20130101) |
Current International
Class: |
G01N
27/00 (20060101); G01N 15/00 (20060101) |
Field of
Search: |
;324/71.4,71.3,71.1,691-693,713,439,450
;73/61.71,61.73,865.5,861.41 ;702/26,29 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
Zhang, Zheng, et al., "An electronic pollen detection method using
Coulter counting principle", Atomospheric Environment 39 (2005)
5446-5453. cited by other .
Jagtiani, Ashish, et al., "Detection and counting of micro-scale
particles and pollen using a multi-aperture Coulter counter",
Measurement Scient and Technology, 17 (2006), 1706-1714; Institute
of Physics Publishing, UK. cited by other .
Jagtiani, Ashish, et al., "A label-free high throughput
resistive-pulse sensor for simultaneous differentiation and
measurement of multiple particle-laden analytes", Journal of
Micromechanics and Microengineering, 16 (2006) 1530-1539; Institute
of Physics Publishing, UK. cited by other .
Paper No. IMECE2006-15540; Zhe, Jiang, et al., "A Microfluidic
Based High Throughput Resistive Pulse Sensor", ASME International
Mechanical Engineering Congress and Exposition, Nov. 5-10, 2006,
Chicago, IL. cited by other.
|
Primary Examiner: Nguyen; Hoai-An D
Attorney, Agent or Firm: Crimaldi; Joseph J. Roetzel &
Andress
Claims
What we claim is:
1. A multichannel particle counting device comprising: a plurality
of microfluidic channels for dividing a first reservoir and a
second reservoir, and for maintaining fluid communication between
the first and second reservoirs; each microfluidic channel
including a control electrode, wherein each control electrode is
substantially electrically isolated from any other control
electrode; each control electrode encoded to respond to a specific
frequency; the first and second reservoirs together including one
set of detection electronics including a first electrode in
electrical communication with a power supply and a second electrode
in electrical communication with a measuring circuit; the one set
of detection electronics being the collector for each signal from
each control electrode and having a means to deconvolute the
collected signals; the reservoirs containing an electrolyte
solution containing particles to be counted; and a means for
creating a net fluid flow of electrolyte from one reservoir to the
other reservoir through the microfluidic channels.
2. The device of claim 1 wherein the control electrode is replaced
by one of a micro-actuator and a nanoactuator.
3. The device of claim 2 wherein the control electrode is replaced
by a microactuator that deforms the microfluidic channel.
4. The device of claim 3 wherein the micro-actuator is a
cantilevered device.
5. The device of claim 2 wherein the control electrode is replaced
by a nanoactuator.
6. The device of claim 5 wherein the nano-actuator comprises nano
polymer brushes.
7. The device of claim 1 wherein the means for creating a net fluid
flow is the presence of pressure, an electro-osmotic state and an
electrophoretic state.
8. The device of claim 1 wherein the microfluidic channels are
formed by at least one of lithography and micro-machining.
9. The device of claim 1 wherein the electrodes comprising the one
set of detection electrodes are selected from Ag/AgCl, platinum,
graphite, or a combination thereof.
10. The device of claim 1 wherein each control electrode comprises
at least one of Ag/AgCl, platinum, or graphite.
11. A method for rapidly counting particles comprising: charging
one reservoir of the device of claim 1 with an electrolyte solution
containing at least one particle to be measured; applying a voltage
across the electrodes comprising the one set of detection
electrodes; allowing the particles to migrate from one reservoir to
the other reservoir through the plurality of microfluidic channels;
dynamically modulating the microfluidic channels; the one set of
detection electrodes detecting multiple signals generated as
particles pass through the plurality of microfluidic channels;
deconvoluting the signals detected; correlating the signals to the
number of particles passing through each microfluidic channel; and
counting the deconvoluted signals.
12. The method of claim 11 wherein the signals are detected
simultaneously and collected simultaneously by the one set of
detection electronics.
13. The method of claim 11 wherein the microfluidic channels are
dynamically modulated by a control electrode in each microfluidic
channel.
14. The method of claim 13 wherein each control electrode is
encoded to a different signal frequency.
15. The method of claim 11 wherein the microfluidic channels are
dynamically modulated by at least one of a micro- or
nano-actuator.
16. The method of claim 15 wherein each channel contains a separate
micro- or nano-actuator.
17. The method of claim 15 wherein the micro- or nano-actuator is a
channel deforming device.
18. The method of claim 17 wherein the deforming device responds to
an electrical impulse by cantilevering the micro-actuator.
19. The method of claim 15 wherein the nano-actuator comprises nano
polymer brushes that extend or relax in response to electric charge
present within the device.
20. The method of claim 11 wherein the particles being counted are
selected from one or more of pollen, dust, airborne contaminants,
microbes, viruses, and biological warfare agents.
21. The method of claim 11 wherein the applied voltage is between 1
and 4 volts.
22. The method of claim 11 wherein the signals comprise current
and/or voltage pulses.
23. The method of claim 11 wherein the combined signals are
deconvoluted by applying the Hardmard Transformation or Fast
Fourier Transformation technique to the signal data.
Description
BACKGROUND OF THE INVENTION
The present invention is generally related to a multichannel
particle counting method and a device for practicing the method.
Such counters can be used to count micro-scale and/or nano-scale
particles and the like. Counters within the scope of the present
invention generally operate by sensing changes in resistance,
conductivity, conductance or the like. More particularly, as a
particle passes through a channel, it disrupts the ion current
therein, thus increasing the channel's resistance.
Quantitative measurements of the size and concentration of micro
and nano scale particles has been accomplished using Coulter
counters. A typical Coulter counter device comprising a single
micropore that separates two chambers containing electrolyte
solutions. When a particle flows through the microchannel, it
results in the electrical resistance change of the liquid filled
microchannel. The resistance change can be recorded in terms of
current or voltage pulses, which can be correlated to size,
mobility, surface charge and concentration of the particles. Due to
the simple construction of these devices and the reliable sensing
method, Coulter devices have found application in a broad range of
particle analyses from blood cells to polymeric beads, DNA, virus
particles and even metal ions.
One substantial disadvantage of existing Coulter counters is their
low throughput efficiency, which substantially extends measurement
times. Coulter counting measurement relies on particles passing
through a tiny orifice (microchannel) one by one from one chamber
to the other. Thus, in order to complete sampling of a small number
of particle solutions, thousands of micro or nanoparticles have to
pass through the orifice one by one, which could be prohibitively
time consuming. For instance, one estimate shows that a sample
having a particle concentration of 10.sup.8 particles/mL (v/v ratio
0.026%) requires 27.7 hours to complete a measurement, assuming
each particle takes about 0.05 seconds to pass through the orifice,
only one particle is resident in the orifice at any given time, and
assuming a 0.01 mL sample volume. The measurement time is further
extended as the orifice size decreases.
A variety of approaches to alleviating the time-measurement issue
have been tried in the art. For instance, electroosmosis and
electrophoresis have been applied to drive particles and
electrolyte fluids. However, both methods have fallen short.
Particularly, in order to obtain a sufficient fluid velocity, a
strong external electric field must be applied leading to high
power consumption, which is not practical for most biological
applications. Furthermore, electroosmosis and electrophoresis only
drive charged particles. Thus, if the particles are only slightly
charged or neutral, electric forces are too weak to substantially
shorten measurement time. Accordingly, there is a deficiency in the
art in that it lacks a high throughput particle counting method and
device, which is compatible with biological particles.
The present invention overcomes the challenges and deficiencies of
the prior art by providing a particle counting method and device
having a plurality of orifices, which are capable of counting
particles in parallel with one another. Furthermore, such systems
are compatible with biological particles inasmuch as it circumvents
the need for electrophoretic or electroosmotic fields. Thus, the
present invention fills a substantial gap in the art.
SUMMARY OF THE INVENTION
The present invention is generally directed to a multichannel
particle counting device comprising a plurality of microfluidic
channels dividing a first reservoir and a second reservoir and
maintaining fluid communication therethrough; each microfluidic
channel including a control electrode, wherein each control
electrode is substantially electrically isolated from every other
control electrode; and each control electrode encoded to respond to
a specific frequency; the first and second reservoirs including one
set of detection electronics including a first electrode in
electrical communication with a power supply and a second electrode
in electrical communication with a measuring circuit; the one set
of detection electronics being the collector for each signal from
each control electrode and having a means to deconvolute the
collected signals; the reservoirs containing an electrolyte
solution containing particles to be counted; and a means for
creating a net fluid flow of electrolyte from one reservoir to the
other reservoir through the microfluidic channels.
A method for rapidly counting particles comprising the steps of
charging one reservoir of the foregoing device with an electrolyte
solution containing at least one particle to be measured; applying
a voltage across the one set of detection electronics; allowing the
particles to migrate from one reservoir to the other through the
plurality of microfluidic channels; dynamically modulating the
microfluidic channels; the one set of detection electronics
detecting the signals generated as particles pass through the
plurality of microfluidic channels; deconvoluting the signals
detected; correlating the signals to the number of particles
passing through each microfluidic channel; and counting the
deconvoluted signals.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic showing the a multichannel particle counting
device;
FIG. 2 is a diagram pictorially showing the resistance of each
sensing element. Note that R and R' (i.e., the resistances of the
electrolytes outside the channel) can be neglected when compared to
the channel resistances R.sub.ch and R'.sub.ch;
FIG. 3 is an electrical schematic showing an example embodiment of
the present invention comprising a 3-channel multiplexing particle
counting device;
FIG. 4 is a diagram of a airborne pollen sampling device fitted
with the multichannel particle counting device of the present
invention;
FIG. 5 is a drawing of a multi-aperture embodiment for
microparticle detection;
FIG. 6 (a) is a schematic front view of a single channel of a
multi-aperture embodiment, (b) is a magnified view of a single
channel; and (c) is a drawing of an circuit equivalent of a single
channel;
FIG. 7 is a equivalent electric circuit of a four-aperture
embodiment;
FIG. 8 is a set of photomicrographs of (a) 20 .mu.m
polymethacrylate particles, (b) 40 .mu.m polymethacrylate
particles, (c) Juniper Scopulorum pollen, and (d) Cottonwood
pollen;
FIG. 9 is a set of four voltage traces obtained from four different
sampling resistors in response to 40 .mu.m PM particles;
FIG. 10 is a graph showing typical results of a cross-correlation
analysis performed on 40 .mu.m PM particles;
FIG. 11 is a set of four plots showing relative resistance of four
sensing microapertures as a function of time, wherein each pulse
corresponds to a single particle;
FIG. 12 is a typical resistive pulse resulting from a 40 .mu.m PM
particle passing through a microaperture;
FIG. 13 is a set of four histograms showing particle size data
obtained from each of four channels of the four-channel
embodiment;
FIG. 14 is a set of four voltage traces obtained from each of the
four sampling resistors and shows the voltage response due to
Juniper pollen particles;
FIG. 15 is a set of four plots of the relative resistance due to
Juniper pollen particles;
FIG. 16 is a typical resistive pulse resulting from a Juniper
pollen particle;
FIG. 17 is a pair of plots showing the relative resistance pulses
from a single channel embodiment due to (a) 20 .mu.m PM particles,
and (b) cottonwood pollen particles;
FIG. 18 is a pair of schematics showing (a) a front view of a
single channel embodiment; and (b) a magnified view of a single
channel embodiment;
FIG. 19 is a pair of electrical equivalent circuits of (a) a single
channel embodiment, and (b) a four channel embodiment;
FIG. 20 is a set of three photomicrographs of (a) 20 .mu.m PM
particles and 40 .mu.m PM particles, (b) cottonwood pollen, and (c)
Juniper Scopulorum pollen;
FIG. 21 is a set of plots showing (a) data from four sampling
resistors, and (b) magnified voltage pulses;
FIG. 22 is a set of plots showing (a) the relative resistance of
each channel, and (b) magnified resistive pulses for each
channel;
FIG. 23 is a pair of histograms showing the estimated particle size
of (a) a 40 .mu.m particle, and (b) a 20 .mu.m particle;
FIG. 24 is a set of drawings showing (a) a qualitative illustration
of resistive pulse shape, (b) the result of a neutral particle
entering a channel, and (c) the result of a charged particle
entering a channel;
FIG. 25 is a scatter plot of the relative resistive pulse heights
due to four different particles, i.e., 20 .mu.m PM, 40 .mu.m PM,
cottonwood pollen, and Juniper pollen;
FIG. 26 is a plot of relative resistance as a function of time for
(a) channel 1, and (b) channel 2;
FIG. 27 is a schematic drawing of a micromachined multichannel
resistive pulse sensor embodiment;
FIG. 28 is an electrical schematic showing a simplified model of a
multichannel resistive pulse sensor embodiment;
FIG. 29 is a typical crosstalk analysis between adjacent
microchannel;
FIG. 30 is a pair of photomicrographs showing (a) 40 .mu.m PM
particles, and (b) Juniper Scopulorum pollen;
FIG. 31 is a set of plots showing typical relative resistance
traces from four channels;
FIG. 32 is a typical cross correlation analysis for adjacent
channels;
FIG. 33 is a typical relative resistance for a mixture of Juniper
pollen and 40 .mu.m PM particles in a one-channel embodiment;
FIG. 34 is a pair of magnified resistive pulses due to (a) Juniper
pollen, and (b) 40 .mu.m PM particles;
FIG. 35 is a graphical summary of the measured diameter and
concentration of 40 .mu.m PM particles in a mixture;
FIG. 36 is a mail sorting embodiment for detecting Anthrax spores
on mail items;
FIG. 37 is a fluid dynamics simulation of the behavior of (a) 1 mm
and (b) 1 .mu.m particles in a fluid flow having a 10 m/s inlet
velocity;
FIG. 38 is a schematic view of a microfluidic chip and a
microactuator chip embodiment before bonding;
FIG. 39(a) is a schematic view of an actuator embodiment comprising
PMN-PT/Si bimorph and pin element bonded to a microchannel;
FIGS. 39(b) and 39(c) are schematic views of the actuator
embodiment of FIG. 39(a), wherein the actuator is shown to modulate
flow through the fluid microchannel;
FIG. 40 is a graph of the actuation stroke of a 500 .mu.m long
bimorph cantilever versus applied voltage, wherein "k" is stiffness
and "f" is the first order natural frequency;
FIG. 41(a)-(b) is a diagram of polymer brushes;
FIG. 42(a)-(c) provides a side view of grafted polymer brushes in
the (a) equilibrium state, (b) collapsed state and (c) erect
state;
FIG. 43 provides a graph plotting average polymer brush height
against change in electric field as a function of time;
FIG. 44 provides a PM-FT-IRRAS analysis of the hydrocarbon region
of a polymer brush grown on Au surface;
FIG. 45 is a schematic view of the multiplexed multi-channel design
allowing measurement using single detection electronics;
FIG. 46 is a simplified electrical model for a four channel device
according to FIG. 45;
FIG. 47 is a block diagram for multiplexing using amplitude
modulation for multi-channel resistive pulse sensing;
FIG. 48 (a)-(b) provide graph data for testing results for 30 .mu.m
polystyrene particles passing through a single channel counter;
FIG. 49 provides graphs of measurements for two channels of a four
channel device (a) channel 1 with an encoding frequency of
f.sub.c=20 kHz and (b) channel 4 with an encoding frequency of
f.sub.c=65 kHz;
FIG. 50 provides graphs of (a) a section of the filtered signal
from FIG. 49 showing the variation in voltage as a particle passes
through the channel with superimposed output of the envelope
detector and (b) output of the envelope detection showing a voltage
pulse for channel 1;
FIG. 51 provides a graph of the cross correlation analysis for the
signals of two channels from FIG. 48;
FIG. 52 provides cyclovoltammetry (CV) measurement results for the
polymer brushes according to the invention;
FIG. 53 provides further cyclovoltammetry results under specified
conditions and after acetonitrile treatment; and
FIG. 54 provides comparative data for treatment with acetonitrile
at various concentrations.
DETAILED DESCRIPTION OF THE INVENTION
The present invention generally relates to a method for rapidly
counting particles using a plurality of orifices for simultaneously
sensing particles. The present invention also generally relates to
a device for practicing the method of the present invention.
The method of the present invention includes providing a plurality
of orifices that are capable of passing particles to be counted,
wherein the diameter of the orifices is such that they can pass the
particles one at a time, i.e., in single file. In general the
orifices separate two electrolyte solutions, wherein one solution
is in electrical communication with a cathode and the other is in
electrical communication with an anode. When a voltage is applied
across the cathode/anode pair, an ion current flows through the
orifices. Thus, a signal is generated when at least one particle
enters at least one orifice, thereby obstructing the flow of ion
current and raising resistance. The signal can be read conveniently
in terms of current or voltage. Furthermore, the present invention
simultaneously detects particles in a plurality of orifices. Since
these orifices are in a parallel electrical relationship, the
signals generated thereby are multiplexed, and thus must be
deconvoluted. The "Hardmard/Fourier Transformation" makes it
possible to deconvolute the signal of the multiplexed particle
counting device of the present invention.
Membranes within the scope of the present invention can be
fabricated from a wide variety of materials including without
limitation organic polymers such as polymethyl methacrylates,
polycarbonates, polyimides, polyphenols, chlorinated polyolefins,
and the like. Additionally, membranes within the scope of the
present invention can be fabricated from silicon, n-type silicon,
p-type silicon, and the like. Membrane materials within the scope
of the present invention should be stable under ordinary usage
conditions, and should be capable of forming the pores and other
micro and/or nano structures comprising the present invention.
Electrodes within the scope of the present invention can be
fabricated from any of a variety of materials including without
limitation, Ag|AgCl, platinum, and graphite electrodes.
Any of a variety of electrolytes can be used as the electrolyte of
the present invention. In general, acceptable electrolytes are
compatible with the selected electrode(s), and comprise cations and
anions having similar mobilities. For instance, when the selected
electrode is Ag|AgCl acceptable electrolytes include, without
limitation, KCl and NaCl.
An example of the present invention is shown in FIG. 1. Four
microchannels 110 are formed using insulating blocks 112 (i.e.,
isolation blocks). Each individual microchannel 114 (i.e., orifice)
includes a control electrode 124, which carries out control and
measurement functions the nature of which will become apparent in
the following paragraphs. Furthermore, each control electrode 124
is insulated from every other control electrode 124 so that sensing
events occurring in one microchannel 114 do not affect sensing
events in other microchannels 114. The plurality of microchannels
110 separate two electrolyte solutions contained in separate
reservoirs 130, 132. One reservoir 130 is in electrical contact
with a cathode 120 and the other is in electrical contact with an
anode 122. When a suitable voltage (e.g., 2.7 V) is applied across
the electrodes an ion current is generated, which runs through the
plurality of microchannels 110. In this embodiment, either the
cathode 120 or anode 122 is in electrical communication with a
power supply, while the other electrode is in electrical
communication with a measurement circuit. Signals are generated
when one or more particles enter one or more microchannels 114
thereby raising the resistance to ion current therein and causing a
consequent increase in voltage. Each microchannel causes its own
voltage and/or current signal, which superpositions with the
signals of each other microchannel.
FIG. 2 shows the equivalent resistance model of an individual
channel. The resistances R and R' are that of the fluids in the
reservoirs between the cathode (or anode) and the channel. R.sub.ch
is the channel resistance and its value is a function of channel
diameter, length and ion solution in the channel. When a particle
enters the channel 214 it displaces some ions and from the channel
214, which results in an increase in the channel resistance. The
control electrode 224 is positioned in the middle so that the
channel resistance is split into two resistances R.sub.ch,
R'.sub.ch. Thus, the control electrode 224 forms a node with the
two resistances.
In order to detect a particle passing through a particular channel
the response of each individual channel needs to be obtained in the
form of a current or voltage pulse. However, since all of the
channels are in electrical communication with the electrolyte
solutions the signal sensed by the measurement circuit is the sum
of the signals from all channels at any given time. The present
invention is able to deconvolute the raw superposition of signals,
and records the signals of each individual channel. FIG. 3 is a
schematic of a measurement circuit for a 3-channel device. V.sub.cc
is kept at a high voltage level, such as from about 1 V to about 4
V (e.g., 2.7 V). Each control electrode 324 is connected to a
transistor (T1, T2, or T3). The measurement circuit has a
transistor (T4) in common emitter configuration. T4 is always kept
on, so that the voltage at node 4 is kept at approximately 0.7 V.
If T1 is on, the voltage at node 1 is 0.7 V, so that there is no
current input to transistor 4 from channel 1. Because of the
measurement configuration, the output of the transistor T1 is
representative of the resistance of channel 1 (R.sub.1L).
Similarly, if T2 and T3 are open, we could measure the output of T2
and T3, which are representative of the resistances of channel 2
and 3 (R.sub.2L, R.sub.3L) respectively. On the other hand, if T1
is off then current i.sub.1, which is representative of resistance
(R.sub.1L), will be one input to T4 or the sum of any channels
selected. Similarly, if T2 and T3 are off then currents i.sub.2 and
i.sub.3 will be the input to transistor T4. Therefore, by
controlling the on/off state of control electrodes it is possible
to turn the current on and off from each channel and measure the
sum of the responses of any selected channels.
The raw signal obtained from the measurement circuit is
deconvoluted according to the following process. The on/off states
of the four control electrodes (S1, S2, S3) are controlled with a
pseudorandom sequence. The current i4, measured from transistor 4,
is the sum of the current through selected channels. A channel is
selected when its transistor is off. Thus, if a control electrode
is off the current through the channel will be input Transistor
4.
TABLE-US-00001 TABLE 1 S1 S2 S3 I4 OFF OFF OFF i1 + i2 + i3 OFF OFF
ON i1 + i2 ON OFF OFF i2 + i3
Desired current combinations can be measured (Matrix Y) by setting
the desired switching sequence (matrix S) of control electrodes.
For the sequence code in Table 1, the sequence matrix S can be
written as:
##EQU00001##
The Hardmard transformation can be used to find the current
response of individual channels since the switching sequence is
known. Therefore the current response of individual channel X, can
be calculated as the dot product of matrix Y and the inverse of
matrix S: X=S.sup.-1Y Furthermore, four combinations are needed in
order to solve this equation because it entails four unknown
currents.
In a Hardmard Transformation, the mean square error is reduced by a
factor of (n+1).sup.2/4n indicating that the signal-to-noise ratio
is increased by a factor of (n+1)/4n.sup.1/2. Thus, as the number
of channels increases, the signal-to-noise channel improves.
One embodiment of the present invention comprises a device 400 for
quantitatively detecting the concentration and particle size of
pollen in air. For instance, the multiplexed particle counting
device 400 of the present invention can be outfitted with an air
sampling device 410 according to FIG. 4. As shown, a sampling
bottle 412 has an air sampling port 414, which is vented to the
atmosphere. Furthermore, it is fitted with an electrolyte intake
port 420, and fed there through by an electrolyte reservoir 422.
Additionally, the sample bottle has two output ports 430, 440. One
is a vapor line 440, which is in fluid communication with a vacuum
pump 442, so that when the vacuum pump 442 operates gas is drawn
from the sample bottle 412. The other output port is a liquid
output 430, which carries electrolyte solution to the particle
counting portion 432 of the device 400. Optionally, the airborne
pollen sampling device can additionally include a component 434 for
adding antibodies to the electrolyte solution before it reaches the
particle counting portion 432 of the device 400.
This system 400 operates as follows. The vacuum pump 432 draws air
into the sample bottle 412 through air intake port 414. Any
particles that may be present impact the liquid electrolyte surface
416 and are deposited therein while the gas is drawn out of the
sample bottle 412 through the vacuum pump 442. The particle-laden
electrolyte solution then travels through the electrolyte output
port 430 and down the liquid line leading to the particle counting
portion 432 of the device 400, where the particles are then
counted. Furthermore, the liquid can be induced to flow through the
counting portion 432 by maintaining a positive pressure on the
sampling side of the counting portion 432. For instance, as shown
in FIG. 4, an electrolyte reservoir 422 is included, which can be
elevated so that a gravity-induced pressure gradient is created,
which drives liquid flow.
Another embodiment of the present invention comprises a device for
detecting the concentration and particle size of chemical and/or
biological warfare agents such as weaponized (i.e., aerosolized)
anthrax. In still another embodiment, the present invention
comprises a device for detecting toxins, impurities, or microbial
or viral contaminants in waters, such as drinking water. Still
another embodiment of the present invention comprises a device for
rapidly counting blood cells, and/or comparing the number of red
blood cells to white blood cells.
I. Pollen Detection Embodiments:
One non-limiting embodiment of the present invention comprises a
multi-aperture Coulter counter, as shown in FIG. 5. This embodiment
comprises four peripheral reservoirs and a central reservoir. Each
peripheral reservoir is connected to the central reservoir through
a miniature channel. According to this embodiment, a micro-scale
aperture in the middle of each mini-channel is used for
sensing.
In this embodiment, the central reservoir and one half of each of
the four mini-channels can be formed by drilling holes in a
polymethyl methacrylate (PM) block. Furthermore, each of four
additional PM blocks can be drilled to form the other half of a
mini-channel and a peripheral reservoir. After the holes are
drilled, the PM blocks can be cleaned, for example, with ethyl
alcohol and/or sonicated in an ultrasound bath. The microapertures
are fabricated by piercing four polymer membranes with a heated
micro needle. The membranes can be examined under a high-precision
microscope, and the microapertures are found to have diameters
between 90 .mu.m and 110 .mu.m, as shown in Table 2.
TABLE-US-00002 TABLE 2 Aperture 1 Aperture 2 Aperture 3 Aperture 4
Diameter D 110 .mu.m 110 .mu.m 90 .mu.m 100 .mu.m
The nominal thickness of the membrane is about 100 .mu.m. However,
due to the hot-piercing microaperture fabrication process, the
length of each microaperture might change considerably from the
nominal thickness of the membrane, which is determined later.
According to this non-limiting embodiment, the first channel can be
formed by applying epoxy to one mini-channel side of the PM block
with the central reservoir, and to the mini-channel side of one of
the PM blocks with a peripheral reservoir. A membrane can be placed
between the two blocks, carefully aligned so that its microaperture
is centered between the two halves of the mini-channel. The blocks
are then clamped together for two to five hours, or until the two
blocks and the membrane are firmly attached. A pair of 1 mm holes,
located 5 mm away from the membrane on both sides, can be drilled
in the PM blocks. The Ag/AgCl electrodes are placed on both sides
of the membrane through the 1 mm holes. Then epoxy can be applied
to fix the electrodes and seal the mini-channel. The same procedure
is repeated for the other three peripheral blocks to form a
four-aperture sensor. One of ordinary skill in the art will readily
appreciate that a variety of alternative materials, fabrication
techniques, and electrodes can be used to form a device within the
scope of the present invention.
FIG. 6(a) shows the sectioned front view of a single sensing
channel (across the A-A line in FIG. 5), along with the measurement
setup. R.sub.s is a known external sampling resistor. The Ag/AgCl
electrodes placed on both sides of the membrane are used to apply a
constant DC voltage VCC across the channel. FIG. 6(b) shows a
magnified drawing of the mini-channel, microaperture and
electrodes. The measurement architecture of one sensing channel is
electrically equivalent to the circuit in FIG. 6(c), where R.sub.c
is the resistance of the electrolyte-filled microaperture.
As a particle passes through the microaperture there is a change in
the electrical resistance of the aperture. This leads to a change
in the voltage VS across the measurement resistance R.sub.s. From
the circuit model in FIG. 6(c), the relative change in the
resistance of the microaperture is given by:
.delta..times..times.'.times..times.' ##EQU00002## where
.delta.R.sub.c is the change in aperture resistance, R.sub.c is the
resistance of the aperture when no particles are present, V.sub.CC
is the applied DC voltage, V.sub.s is the voltage measured across
the sampling resistor when the aperture is filled only with
electrolyte solution and V.sub.s' is the peak voltage measured
across the sampling resistor as a particle passes through the
microaperture.
For a microaperture with length L and diameter D (see FIG. 6(b)),
the change in resistance as a particle passes through it is given
by:
.delta..times..times.'.times..times..times. ##EQU00003## where d is
the diameter of the particle, L' is the corrected aperture length
to account for fabrication artifacts, which equals L+0.8 D.
Equation 2 holds when (d/D).sup.3<0.1, as is the case in this
and other embodiments. Thus, the particle diameter can be
calculated from the relative change in resistance according to:
.delta..times..times..times.'.times..times..times. ##EQU00004##
In one embodiment, a four-aperture sensor has four sampling
resistors R.sub.s1, R.sub.s2, R.sub.s3 and R.sub.s4 across which
four voltage measurements V.sub.s1, V.sub.s2, V.sub.s3 and V.sub.s4
are made. The overall measurement setup for the four-aperture
sensor is electrically equivalent to the circuit shown in FIG. 7,
where R.sub.c1, R.sub.c2, R.sub.c3 and R.sub.c4 are the resistances
of the four microapertures. According to some embodiments, the four
negative electrodes are electrically shorted in the central
reservoir to ensure that each channel sees the same constant DC
voltage V.sub.CC. In this way, a variation in resistance in one
microaperture does not cause voltage variation (i.e., crosstalk) in
any other channel.
Example of Pollen Detection Embodiment:
One working example of the present invention is set forth as
follows. Four types of micro-scale particles are chosen for a test
of a multi-aperture sensor embodiment. They are polymethacrylate
(PM) particles with diameters of 40 .mu.m and 20 .mu.m, Rocky
Mountain Juniper (Juniper Scopulorum) pollen, and Cottonwood
pollen. All particles are obtained from Sigma Aldrich, Inc. PM
particles are chosen because they are commercially available and
have well-characterized properties. The diameters of the pollen
particles are determined using high-resolution optical microscopy,
and range from 17.5 .mu.m to 22.5 .mu.m for the Juniper pollen and
20 .mu.m for Cottonwood pollen. FIG. 8 shows photomicrographs of
the four types of particles.
For experiments involving the polymethacrylate (PM) particles, 40
.mu.m and 20 .mu.m particle solutions can be prepared by diluting
0.1 mL of the original solution, which has 10% solid content, in 2
mL and 10 mL of deionized water, respectively. The estimated
particle concentrations of the 40 .mu.m and 20 .mu.m particle
solutions are approximately 1.2.times.10.sup.5 mL.sup.-1 and
2.times.10.sup.5 mL.sup.-1, respectively. For experiments involving
Rocky Mountain Juniper pollen particles, a solution can be prepared
by diluting 0.1 mL of the original pollen particles in 7 mL of
deionized water.
In each example, 1 mL of the prepared particle solution is added to
the central reservoir using a microsyringe. The liquid in the
central reservoir is agitated to make sure that the particles are
well dispersed. A gravity-induced pressure difference is created by
placing the central reservoir at a higher level than the peripheral
reservoirs. Pressure-driven flow forces the particle solution to
move towards the peripheral reservoirs through the four sensing
apertures.
The sampling resistor for each channel is R.sub.s=100 k.OMEGA., and
the applied voltage across the electrodes of each channel is
V.sub.CC=3V. The entire measurement architecture is placed in a
Faraday cage to reduce and/or control noise. As the particles pass
through the microapertures, voltage pulses across all sampling
resistors can be recorded simultaneously using, for example, a
National Instruments NI-6220 data acquisition board. The voltages
can be monitored in real time using, for example, LabView software
with a sampling frequency of 20 kHz. The data obtained are
converted to relative resistance change (.delta.R.sub.c/R.sub.c)
using Equation (1). This relative change is used to estimate the
particle diameter (i.e., using Equation (3)). Particle
concentration can be estimated by counting the number of resistive
pulses during a selected time period.
Typical voltage traces resulting from the foregoing example are
shown in FIG. 9. Pulses are recorded during a selected time period
of one second. A typical result of a cross-correlation analysis
performed on the signals from a pair of channels is shown in FIG.
10. The cross-correlation coefficients between channels are less
than 5%, indicating negligible correlation among the pulses in
different channels. This indicates that the four sensing apertures
are able to simultaneously generate voltage pulses and count
particles with negligible crosstalk among channels. Notably,
different channels have different base voltages (V.sub.s). This is
primarily because the base resistances of the four microapertures
(R.sub.c1, R.sub.c2, R.sub.c3 and R.sub.c4) differ due to
fabrication artifacts.
The ratio of the resistance change (.delta.R.sub.c/R.sub.c) for
each microaperture, calculated using Equation (1), is plotted as a
function of time in FIG. 11. FIG. 12 shows a more detailed view of
a typical pulse for a 40 .mu.m polymethacrylate particle from FIG.
11. It can be seen that the duration of the pulse due to the
particle passing through the aperture is about 2 ms. Hence the
average speed of a particle traveling through the channel is
approximately 0.05 ms.sup.-1. This is about the same as the
velocity of the fluid flow. The average relative change in
resistance is used to calibrate the length of each aperture using
Equation (2). This calculation assumes that the aperture diameter
measurement is accurate. The calibration results are shown in Table
3.
TABLE-US-00003 TABLE 3 Aperture 1 Aperture 2 Aperture 3 Aperture 4
Calibrated 173 .mu.m 153 .mu.m 118 .mu.m 127 .mu.m Length (L)
The calibrated aperture length can be used to calculate particle
diameter using Equation (3). FIG. 13 shows a histogram of estimated
particle size, along with the average size, standard deviation
(.sigma.) and number of particles (n) for each channel. The
estimated particle diameters lie between about 37.28 .mu.m and
43.25 .mu.m (39.94.+-.1.41 .mu.m) for channel 1, between about
37.73 .mu.m and 42.06 .mu.m (39.97.+-.1.04 .mu.m) for channel 2,
between about 37.78 .mu.m and 41.9 .mu.m (40.+-.1.22 .mu.m) for
channel 3 and between about 36.44 .mu.m and 44.92 .mu.m
(39.92.+-.1.96 .mu.m) for channel 4. The manufacturer specifies the
actual diameter of the particles to be 40.+-.0.8 .mu.m. The
measurement error in particle size is approximately within the
overall uncertainty error range. The differences are likely due to
the uncertainty of the microaperture dimension, electronic noise
and the off-axis position when a particle passes through the
microaperture.
The concentration of the particles in the four channels can be
calculated from the number of peaks during a one second period as
shown in FIG. 11. According to this example, the concentrations are
found to be 1.09.times.10.sup.5 mL.sup.-1, 0.95.times.10.sup.5
mL.sup.-1, 1.04.times.10.sup.5 mL.sup.-1 and 1.12.times.10.sup.5
mL.sup.-1 for channels 1, 2, 3 and 4, respectively. The measured
particle concentration in each channel is slightly lower than the
estimated particle concentration, which is about 1.2.times.10.sup.5
mL.sup.-1. This is possibly because some PM particles are deposited
onto the substrate during the experiments.
Typical voltage traces with pulses are shown in FIG. 14, and are
recorded over a time period of one second. It can be observed that
the polarity of the voltage pulses when a pollen particle passes
through the microaperture is opposite to that caused by a
polymethacrylate particle (FIG. 9). As is the case for the PM
particle experiments, the cross-correlation analysis shows
negligible crosstalk among channels. The measured voltages are
converted to relative changes in aperture resistance as shown in
FIG. 15. The downward resistance pulse corresponds to a decrease in
the resistance of the microaperture when a pollen particle passes
through the aperture. A typical pulse for a Juniper pollen particle
taken from FIG. 15 is shown in FIG. 16.
This phenomenon indicates that a particle affects the microaperture
resistance in two competing ways. First, it displaces electrolyte
solution in the microaperture, thereby reducing the number of free
ions inside the microaperture, which leads to an increase in
resistance. Second, if it has a surface charge, it brings
additional charges into the microaperture, which leads to a
decrease in resistance. According to the results from this example,
the pollen particles have high surface charge, while the PM
particles are only slightly charged. When the surface charge is
high and the concentration of ions in the electrolyte solution is
low, as is the case for pollen particles in this example, the
second factor is dominant, and the overall effect of a pollen
particle passing through a microaperture is a downward resistive
pulse. This phenomenon can be used to differentiate pollen
particles from other only slightly charged particles. It is also
possible to measure pollen particle size using electrolyte solution
of high concentration, so that the particle size plays the dominant
role in the size of the resistive pulse.
In order to further demonstrate that this embodiment can be used to
differentiate various particles, two additional particles, 20 .mu.m
polymethacrylate particles and Cottonwood pollen, are tested using
a single Coulter cell (channel 1 with the 110 .mu.m aperture).
These two particles are chosen because they are similar in size to
Juniper pollen but may differ in surface properties.
Typical traces of resistive pulses are shown in FIGS. 17(a) and
(b). These traces show resistive pulses that are calculated from
voltage signals recorded over a one second period. The traces
correspond to the 20 .mu.m PM particles, and the Cottonwood pollen.
The particle diameters are calculated from the resistive pulse data
shown in FIG. 17(a). Using the calibrated aperture length of 173
.mu.m, the statistical analysis shows that the estimated particle
diameter is 22.46.+-.2.1 .mu.m. The difference between the
calculated and the actual particle diameter (20 .mu.m.+-.0.5 .mu.m)
can be minimized by calibrating both the aperture diameter and
aperture length using a number of quasi-monodisperse particles of
standard sizes. While the estimated particle diameters appear to
have a larger divergence about the average than that specified by
the manufacturer (20.+-.0.5 .mu.m), the measurement error is
approximately within the overall uncertainty error range. The
concentration of the 20 .mu.m PM particles is calculated from the
number of peaks during a period of one second. According to this
example, the concentration is calculated to be 1.91.times.10.sup.5
mL.sup.-1, compared to the original concentration estimate of
2.times.10.sup.5 mL.sup.-1.
Like Juniper pollen, Cottonwood pollen particles generate downward
resistive pulses (a decrease in resistance) when they pass through
the microaperture (FIG. 17(d)). However, the resistive pulse height
(relative to the base resistance of microaperture) generated by
Cottonwood pollen (0.478%.+-.0.226%) is considerably lower than
that of the Juniper pollen (2.73%.+-.0.99%). While not wishing to
be bound to any one theory, this phenomenon may be attributed to a
difference in surface charge. The influence of aperture geometry
can be eliminated by normalizing the relative resistive pulses for
20 .mu.m and 40 .mu.m PM particles, Juniper pollen, and cottonwood
pollen using the following equation:
.delta..times..times..delta..times..times.'.times..times..times.
##EQU00005##
A scatter plot of normalized .delta.R.sub.c/R.sub.c for the four
particles is shown in FIG. 18. It indicates that the four tested
particles can be identified in a mixture by using the polarity and
magnitude of the resistive pulses. While not wishing to be bound to
any one theory, the large variation in resistive pulses of Juniper
tree pollen may be due to a variation in particle size and shape,
and in surface charge.
The results for both the PM particle and pollen experiments
indicate that this instrument is capable of counting particles
through the four microapertures simultaneously. In contrast to a
single channel Coulter counter, the counting efficiency is improved
by a factor of approximately three. This counting efficiency can be
further improved by integrating more sensing apertures in a
micromachined device. The noise in the sensed voltages averages to
about 0.1 mV. Thus, this embodiment should be able to detect
particles that produce pulses larger than this noise level.
Accordingly, this embodiment is capable of detecting particles with
diameters larger than approximately 8.2 .mu.m, or 6.9% of the
microaperture diameter. We expect that the sensitivity can be
improved by using better shielding and electronics to reduce the
noise level.
The foregoing example demonstrates that some embodiments of the
present invention can be used to distinguish between kinds of
particles and to count pollen and other particles with
significantly improved efficiency.
According to this example, uncertainty analysis is carried out
using the methods of Moffat, Kline and Coleman and Steele. There
are three sources of uncertainty in the estimation of the particle
size. The first source is due to uncertainty in measurement of the
microaperture diameter and in the calibration procedure used to
determine microaperture length. Due to the fabrication process used
to make the microapertures, the microapertures vary in size. The
uncertainty in measuring the diameter is .+-.10%. The calibration
process for determining the length presumes that the measured
diameter is correct. The uncertainty in diameter causes a maximum
uncertainty of .+-.42.5% for microaperture length. Together, the
uncertainties in aperture diameter and length contribute a maximum
of .+-.12.1% uncertainty in particle size evaluation. Note that
this source of uncertainty systematically alters the estimates of
the particle diameters and can be reduced by calibrating the sensor
using quasi-monodisperse particles in a number of standard sizes.
Taking data using two different particle sizes, for example, allows
the practitioner to solve for both the effective diameter and the
effective length of a microaperture, thereby reducing uncertainty
for both parameters.
The second source of uncertainty is due to fluctuations in the
output voltage, which are about .+-.0.05 mV at base voltage levels
of 0.22 V. These fluctuations could be due to either flow
unsteadiness or the measurement electronics, and appear to have no
systematic trend. According to Equation (3), this uncertainty
contributes to an uncertainty of .+-.0.58% and .+-.4.5% in particle
diameter estimation for 40 .mu.m and 20 .mu.m polymethacrylate
particles, respectively.
The third source of uncertainty is due to particles passing
off-axis through the aperture. Given the shape of the pulses
observed, we expect a maximum increase of about 10% in the measured
response. This corresponds to an uncertainty of .+-.3.3% in
particle size. Combining the three uncertainty sources, the
uncertainties of particle size estimation are .+-.12.5% and
.+-.13.2% for 40 .mu.m and 20 .mu.m polymethacrylate particles,
respectively. The foregoing results show that the measurement error
of the counter is well within this uncertainty error range.
II. Label-Free Resistive Pulse Sensor Embodiments:
FIG. 5 shows general structure that is consistent with the
following embodiment. Similar to the previous embodiment, this
embodiment comprises four peripheral reservoirs and a central
reservoir. Each peripheral reservoir is connected to the central
reservoirs through a mini channel. A microchannel, fabricated on a
polymer membrane, is positioned in the middle of each mini channel
and used for particle sensing.
FIG. 18(a) shows the sectioned schematic front view of a single
sensing channel along with the measurement setup. R.sub.s is a
known external sampling resistor. The Ag/AgCl electrodes placed on
both sides of the membrane is used for applying a constant DC
voltage V.sub.cc. FIG. 18(b) shows a blow-up drawing of the
mini-channel, microchannel and electrodes.
The measurement circuit for one sensing channel is equivalent to
the circuit in FIG. 19(a), where R.sub.c is the resistance of the
electrolyte-filled microchannel. R.sub.c has a variation
.delta.R.sub.c when a particle passes through the channel, because
it displaces some of the electrolyte solution in the microchannel.
This causes a change in the measured voltage V.sub.s across the
sampling resistor R.sub.s. From the circuit model in FIG. 19(a),
the relative change in the resistance of the microchannel is given
by Equation (1), where .delta.R.sub.c is the change in channel
resistance when a particle passes through the microchannel, V.sub.s
is the voltage measured across the sampling resistor when the
channel is filled only with electrolyte solution and V.sub.s' is
the peak voltage measured across the sampling resistor as a
particle passes through the microchannel.
For a microchannel with length L and diameter D (see FIG. 18(b)),
the change in resistance as a particle passes through it is given
by Equation (2). Thus, the particle diameter can be calculated from
the relative change in resistance based on Equation (3).
According to one very specific example, the central reservoir and
half of each of the four mini channels can be fabricated by
drilling holes in a polymethyl methacrylate (PM) block. In each of
four additional PM blocks, holes are drilled to form the other half
of a mini channel and a peripheral reservoir. According to this
example, the central reservoir is 12 mm in diameter and 10 mm deep.
Each peripheral reservoir is 10 mm in diameter and 10 mm deep. The
mini channel is 4 mm in diameter. After the holes are drilled, the
PM blocks are cleaned with isopropanol and sonicated in an
ultrasound bath. The microchannels are fabricated by carefully
piercing four polymer membranes with a heated micro needle. The
membranes are inspected under a high precision microscope and the
diameters of the microchannels are measured to be between 120 .mu.m
and 130 .mu.m, as shown in Table 4. The thickness of the membrane
(and therefore the length of each microchannel) is measured to be
approximately 100 .mu.m.
TABLE-US-00004 TABLE 4 Channel 1 Channel 2 Channel 3 Channel 4
Length L 100 .mu.m 100 .mu.m 100 .mu.m 100 .mu.m Diameter D 120
.mu.m 120 .mu.m 130 .mu.m 120 .mu.m
To assemble the device, the PM block with the central reservoir and
one of the PM blocks with a peripheral reservoir are picked, and
epoxy is applied on the mini channel side of the two blocks. A
membrane is placed between the two blocks and is carefully aligned
so that its microchannel is centered between the two halves of the
mini channel. The blocks are then clamped together and kept about
two to five hours, or until the two blocks and the membrane are
firmly attached together. A pair of 1 mm holes, located 5 mm away
from each membrane on both sides, is drilled on the PM blocks. The
1 mm diameter Ag/AgCl electrodes are placed on both sides of the
membrane through the 1 mm holes. Then epoxy is applied to fix the
electrodes and seal the mini channel. The same procedure is
repeated for the other three peripheral blocks to form a
four-channel sensor.
Four microparticles, polymethacrylate particles with
well-characterized diameters of 20 .mu.m (20 .mu.m.+-.0.5 .mu.m)
and 40 .mu.m (40 .mu.m.+-.0.8 .mu.m) (Sigma Aldrich Inc.), Rocky
Mountain Juniper (Juniper Scopulorum) tree pollens (Sigma Aldrich
Inc.) and Populus deltidoes/Eastern Cottonwood pollens (Sigma
Aldrich Inc.) are chosen for testing. These particles are chosen
because they are commercially available and have well-characterized
properties. The diameters of pollen particles are determined using
high resolution optical microscopy. The cotton pollen had a
diameter of about 20 .mu.m. The Juniper tree pollen is egg-shaped
and its diameter ranged from 17 .mu.m to 23 .mu.m. FIG. 20(a) shows
pictures of the 20 .mu.m and 40 .mu.m polymethacrylate particles.
FIGS. 20(b) and (c) shows the pictures of Cottonwood and Juniper
tree pollen respectively taken with the microscope.
Four particle solutions are prepared before the experiments. 40
.mu.m polymethacrylate particle solutions are prepared by diluting
0.1 mL original particle solution, which has 10% solid content, in
2 mL of deionized water. The yield particle concentration of 40
.mu.m solution is approximately 1.2.times.10.sup.5 mL.sup.-1. 20
.mu.m polymethacrylate particle solutions are prepared by diluting
0.1 mL original particle solution (10% solid content) in 7 mL of
deionized water and the yield particle concentration of 20 .mu.m
solution is approximately 2.8.times.10.sup.5 mL.sup.-1. For
Cottonwood and Juniper tree pollen particles, the particle
solutions are formed by diluting 0.1 mL of the original pollen
particle solutions (10% solid content) in 7 mL of deionized
water.
The prepared particle solution is injected into the peripheral
reservoirs separately using a micro syringe. Juniper tree pollen
particles, Cottonwood particles, 40 .mu.m and 20 .mu.m
polymethacrylate particles are loaded into channels 1, 2, 3 and 4
respectively. The liquid in each peripheral reservoir is agitated
to make sure that the particles are well dispersed. A pressure
difference is formed by setting a level difference between the
peripheral reservoirs and central reservoir. The particle solutions
are driven to move from the peripheral reservoirs towards the
central reservoir, which now act as a collecting/sink
reservoir.
The entire measurement setup is placed in a Faraday cage to reduce
noise. The applied electric potential across the pair of electrodes
of each channel is V.sub.cc=3 V. The sampling resistor for each
channel is 100 k.OMEGA.. As the particles passes through the
microchannels, voltage pulses across all sampling resistors are
recorded simultaneously using a National Instruments NI-6220 data
acquisition board. The voltages are monitored in real-time using
LabView software with a sampling frequency of 20 KHz. The data
obtained are converted to relative resistance change
(.delta.R.sub.c/R.sub.c) using Equation (1). The relative change is
used to estimate the particle diameter (using Equation (3)) and
particle concentration.
Example of Label-Free High-Throughput Resistive-Pulse Sensing:
The typical measurement results of voltage traces across the four
sampling resistors during a selected period of time are shown in
FIG. 21(a). A few magnified pulses showing more details of the
pulse shape are shown in FIG. 21(b). It is obvious that the voltage
pulses appear in random sequence. The cross correlation analysis is
performed between the signals from two sensing channels at a time.
We found that the cross correlation coefficients are all less than
5%, indicating there is no correlation among the pulses of
different channels. This implies that the four sensing channels can
simultaneously detect and count particles without crosstalk among
the channels. Note that the difference in the base voltage
(V.sub.s) is due to the base resistance difference (R.sub.c1,
R.sub.c2, R.sub.c3, R.sub.c4) among microchannels because of the
fabrication variation. The particle travel velocity in the
microchannel is estimated by measuring the pulse width (i.e., time
one particle took to pass the microchannel) and the length of the
microchannel, which is used to estimate the particle concentration
later. Some resistive pulses shown in FIGS. 21(b) and 22(b) have a
steeper slope when the particles exit the microchannel than they
enter the microchannel. This may be a result of the particles
entering the microchannel at the center and exiting the
microchannel near the channel wall at an angle.
The voltage pulses of each microchannel are converted to the ratio
of the resistance change using equation. The results are plotted in
FIG. 22(a) as a function of time, along with a few magnified
resistive pulses showing more details in FIG. 22(b). It is obvious
that the four types of particles can be differentiated based on the
direction (downward or upward) and height of resistive pulses.
The relative changes of resistance in channels 3 and 4 are used to
calculate the polymethacrylate particle diameters using Equation
(3). FIG. 23 shows a histogram of the estimated particle size,
along with the average size and standard deviation in size, for
channels 3 and 4. The estimated particle diameters lie in the range
of about 32.89 to 48.96 .mu.m (average 39.81 .mu.m,
.sigma.=.+-.3.41 .mu.m) for channel 3, about 15.68 to 28.18 .mu.m
(average 20.11 .mu.m, .sigma.=.+-.2.91 .mu.m) for channel 2. The
estimated particle size appears to have relatively larger
divergence compared to the actual diameter of the polymethacrylate
particles specified by the manufacturer, which are 40.+-.0.8 .mu.m
and 20.+-.0.5 .mu.m. This is possibly because of the uncertainties
in microchannel dimension, electronic noise and the off-axis
position when particles pass through the microchannel. The
measurement error in the particle size is approximately within the
overall uncertainty error range. From the number of peaks appearing
in channels 3 and 4 during a period of one second, the
concentrations of particles are calculated to be
1.33.times.10.sup.5 mL.sup.-1 (estimated actual concentration is
1.2.times.10.sup.5 mL.sup.-1), 2.46.times.10.sup.5 mL.sup.-1
(estimated actual concentration 2.8.times.10.sup.5 mL.sup.-1),
respectively. The calculated concentrations from measured
resistance pulses are in good agreement with the estimated actual
concentrations. The slight difference is possibly because of the
non-uniformity of particle distribution in the solution. The
results show that the device is capable of counting
polymethacrylate particles and determining their sizes
accurately.
As shown in FIG. 22, in channels 1 and 2, the resistance pulses
caused by Juniper pollens and Cottonwood pollens are all downward.
This implies a decrease in the microchannel resistance when a
pollen particle passed through the microchannel. This phenomenon
can be explained in terms of the surface charge of particles. As
illustrated in FIG. 24(a), a particle affects the ionic current in
two competing ways: first, the particle physically displaces some
of the electrolyte solution and reduces the amount of free ions
inside the microchannel and hence the ion density .sigma.. The
ionic current across the microchannel can be written as:
.intg..times..sigma..mu..times..times..times..times.d ##EQU00006##
where .mu. is the mobility of the free ions, E is the applied
electric field and A is the cross section area. Therefore the
particle induces a decrease in ionic current (.DELTA.I.sub.volume)
as usually expected. Second, if the particle has high surface
charge (see FIG. 24(b)), it induces excess ions in the microchannel
owing to its high surface charge. Hence the ion density .sigma.
increases, leading to an increase in ionic current
(.DELTA.I.sub.charge). When the particle surface charge is high and
the concentration of ions in the electrolyte solution is low, as is
the case in the present example, the ionic current increase
(.DELTA.I.sub.charge) is dominant (FIG. 24(c)). The overall effect
of a particle with high surface charge passing through a
microchannel is an upward ionic current pulse (see FIG. 24(a)).
Therefore, according to Ohm's Law, R=V/l (V is the applied
voltage), a downward resistive pulse will occur. If the surface
charge of a particle is negligible, .DELTA.I.sub.charge is
negligible and an upward resistive pulse is generated. The results
presented in this example suggest that pollen particles are highly
charged, while polymethacrylate particles are slightly charged.
While not wishing to be bound to any one theory, the height of
downward resistive pulses could be explained by the surface charge
of pollen particles. Furthermore, according to this example, the
size of pollen particles can be measured in high concentration
electrolyte solutions (e.g., 0.1 M KCl solution, for instance),
because the induced charge density increase due to pollen is
negligible at concentrations.
As demonstrated by this example, pollen generates downward
resistive pulses. Thus, it can be used to differentiate between
pollen particles from other slightly charged or non-charged
particles. FIG. 25 shows a scatter plot of the relative resistance
.delta.R.sub.c/R.sub.c for the polymethacrylate particles and the
pollen particles during a period of one second. The plot shows that
there are four distinct regions of resistive pulses of the tested
particles. The average resistive-pulse heights and the standard
deviation are calculated and listed in Table 5.
TABLE-US-00005 TABLE 5 Average relative Standard resistive-pulse
height deviation (.delta.R.sub.c/R.sub.c) (%) (%) 20 .mu.m
polymethacrylate 0.23 0.104 40 .mu.m polymethacrylate 1.44 0.392
Juniper pollen -2.83 0.716 Cottonwood pollen -0.478 0.226
According to this example, polymethacrylate particles generate
upward resistive pulses while pollen particles generate downward
pulses. From FIG. 25, the 20 .mu.m polymethacrylate particles and
40 .mu.m polymethacrylate particles can be distinguished by size
exclusion. In spite of the similar sizes of Juniper pollens and
Cottonwood pollens (both are approximately 20 .mu.m), the restive
pulse heights generated by Juniper pollens are approximately six
times higher than that of Cottonwood pollens. While not wishing to
be bound to any one theory, the resistive-pulse difference might be
attributed to the difference in the surface charge of pollen
particles. Therefore, if the polymethacrylate particles and the
pollen particles are mixed in DI water, one can distinguish and
count them separately based on their resistive pulses. FIG. 25 also
shows that the pulse heights of the Juniper tree pollen have more
variations compared to the Cottonwood pollen. One explanation for
this is that the variation is due to particle size variation and
the egg shape of Juniper pollens.
FIG. 25 demonstrates that some embodiments of the present invention
make it is possible to distinguish and count mixtures of various
particles with similar sizes but different surface properties. In
the following example an embodiment is demonstrated to be capable
of distinguishing and counting mixtures of (1) 20 .mu.m
polymethacrylate particles and Juniper tree pollen, and (2) 20
.mu.m polymethacrylate particles and Cottonwood pollen.
Twenty micrometer polymethacrylate particles, Cottonwood pollen and
Juniper pollen solutions are prepared separately as set forth
previously. Mixture 1 is prepared by combining 7 mL 20 .mu.m
polymethacrylate particle solution, and 3 mL Juniper pollen
solution. Mixture 2 is prepared by combining 7 mL 20 .mu.m
polymethacrylate particle solution, and 3 mL Cottonwood pollen
solution. The estimated polymethacrylate particle concentration is
calculated to be 1.99.times.10.sup.5 mL.sup.-1 for both mixtures.
The two particle mixtures are loaded to peripheral reservoirs 1 and
2, respectively. The microchannel diameters for channels 1 and 2
are 100 .mu.m and 110 .mu.m. Channels 3 and 4 are closed using
polymer membranes without microchannels.
Typical traces of resistive pulses converted form the recorded
voltage signal are shown in FIGS. 26(a) and (b). According to this
data, 20 .mu.m PM particles can be differentiated and counted based
on the resistive pulses they generate.
The diameters and concentrations of polymethacrylate particles,
calculated from experimental data in channels 1 and 2, are shown in
Table 6.
Similar to the foregoing results of 20 .mu.m polymethacrylate
particles, the calculated particle diameters have larger variations
than that which is specified by the manufacturer. The calculated
concentrations of 20 .mu.m polymethacrylate particles are
1.72.times.10.sup.5 mL.sup.-1 and 1.94.times.10.sup.5 mL.sup.-1,
compared to the estimated actual concentration of
1.99.times.10.sup.5 mL.sup.-1.
TABLE-US-00006 TABLE 6 Calc. Est. Calculated Vendor's particle
particle particle size specification conc. conc. (.mu.m) (.mu.m)
(mL.sup.-1) (mL.sup.-1) .mu.-channel 1 (100 20.38 .+-. 2.51 20.00
.+-. 0.5 1.72 .times. 10.sup.5 1.99 .times. 10.sup.5 .mu.m) Mixture
1 .mu.-channel 2 (110 20.44 .+-. 2.27 20.00 .+-. 0.5 1 .times.
10.sup.5 1.99 .times. 10.sup.5 .mu.m) Mixture 2
These results indicate that this multi-channel resistive-pulse
sensor is capable of differentiating and counting multiple particle
solutions through the four microchannels simultaneously. In
contrast to a typical Coulter counter that can only analyze one
particle solution, the sensor throughput is improved approximately
300%. The throughput can be further improved by fabricating more
sensing channels in the device. The noise seen in the measured
voltages averaged .+-.0.05 mV, so the device is capable of
detecting particles that produce pulses larger than this noise
level.
This suggests that the device is capable of detecting particles
with diameters larger than approximately 8.9 .mu.m, or 7.4% of the
microchannel diameter. Sensitivity can be improved by improving
shielding, and by the introduction of more sophisticated
electronics to reduce the noise level. The multi-channel sensor
reported herein combines size/surface charge exclusion separation
and high throughput electronic detection in a simple device. The
electrical properties or surface characteristics of biological
particles are of great interest in recent years for novel rapid
assays of these particles. These pollen results indicate that the
multi-channel resistive-pulse device can be used to differentiate
various pollen particles in terms of their surface characteristics
and/or electrical properties. Although only four types of particles
are tested, the resistive pulses due to the passage of various
other biological particles are expected to exhibit distinct signals
because of difference in electrical properties and/or surface
characteristics of biological particles. Thus, some embodiments of
the present invention provide a label-free means for detecting and
counting biological particles. For instance, in addition to the
size/surface charge exclusion, the measurement of the shape of the
resistive pulses provides more detailed information of particles,
including mobility, surface characteristics, electrical properties,
and the like. In one embodiment, this could be done by reducing the
particle travel velocity in the microchannel. Travel velocity can
be controlled by forcing the particles to pass through
microchannels using electrophoresis or a small pressure gradient,
and by using a high sampling frequency.
Because of the simple structure of the multi-channel
resistive-pulse device, throughput can be improved further by
integrating more micro sensing channels. Further, some embodiments
can comprise lab-on-a-chip devices having, for instance
micromachined fluid channels, micro/nano-scale sensing channels and
detection electronics. Additionally, use of multiple sensing
channels enables multiplexing applications. This allows high
throughput signal measurements with a high signal-to-noise ratio
without compromising sensitivity. Therefore, the multi-channel
resistive-pulse sensor embodiment can include a portable, high
throughput micromachined device for micro and nano-scale
bioparticle analysis.
Uncertainty analysis using the methods of Kline, Moffat and Coleman
and Steele is carried out. There are three sources of uncertainty
in the estimation of the particle size. The first source is due to
uncertainty in the measurement of microchannel diameter and length.
Due to the fabrication variation of the microchannel, these
uncertainties are .+-.10% for diameter, and .+-.20% for channel
length. These uncertainties contribute a .+-.10.5% uncertainty in
particle size evaluation. Note that this source of uncertainty
would systematically alter the estimated particle diameters.
The second source of uncertainty is due to the fluctuations in the
output voltage, which are about .+-.0.05 mV at base voltage levels
of 0.2 V. These fluctuations are due to measurement electronics and
appear to have no systematic trend. According to Equation 5, this
uncertainty contributes to an uncertainty of .+-.4.0% and .+-.0.7%
in particle diameter estimation respectively for 20 .mu.m particles
and 40 .mu.m particles.
The third source of uncertainty is due to the off-axis position
when one particle passes through the microchannel. This results in
a maximum uncertainty of about .+-.10% in the resistive pulse,
which corresponds to an uncertainty of .+-.3.2% in particle size.
Combining the three uncertainty sources, the uncertainties of
particle size estimation are .+-.11.7% and 11% for 20 .mu.m
particles and 40 .mu.m particles, respectively. The foregoing
results show that the measurement error of the sensor is
approximately within this uncertainty error range.
III. Microfluidic High-Throughput Resistive Pulse Sensor
Embodiments:
The design concept of one embodiment is illustrated in FIG. 27.
This embodiment comprises a multichannel resistive pulse sensor.
The sensor comprises a single inlet reservoir and a single outlet
reservoir, connected by four microchannels of dimensions 50
.mu.m.times.100 .mu.m.times.300 .mu.m. The device has a common
electrode placed in the inlet reservoir at the entrance of the
microchannels and four central electrodes fabricated at the centers
of the four microchannels. Each central electrode is exposed to the
electrolyte only at the center of the channel for measurement
purposes. The measurement setup for one channel is illustrated in
FIG. 27, and comprises a constant DC power supply V.sub.cc
connected to the common electrode at one end and to a sampling
resistor at the other end. Electrolyte containing particles is
forced to move from the inlet reservoir to the outlet reservoir
through a plurality of sensing channels. When a particle passes
through a channel, it causes a change in the resistance of the
electrolyte-filled channel, thereby resulting in a voltage pulse
across the sampling resistor of that channel. The voltage pulses
across each sampling resistor can be recorded and analyzed
separately. In contrast to a single channel Coulter counter, the
sensor can detect particles through its four sensing channels
simultaneously. Thus, the design enables high throughput.
A simplified electrical circuit equivalent of the measurement setup
is shown in FIG. 28. The measurement electrode in the center of a
sensing channel divides that channel into two equivalent
resistances R.sub.ci and R.sub.ci' (i=1, 2, 3, 4). The first half
of each microchannel (R.sub.ci) serves as a sensing channel, while
the second half of each microchannel serves as an isolation
resistor to reduce or eliminate crosstalk among channels. When a
particle passes through channel i, it affects first the equivalent
resistance of the first half of the channel R.sub.ci, and then the
equivalent resistance of the second half R.sub.ci'. The change is
dependent on both the particle's size and amount of surface charge.
R.sub.si is the sampling resistor of the microchannel, across which
the recorded voltage V.sub.si is measured. R.sub.si is the
resistance formed by the electrolyte between two adjacent
microchannels. This resistance is usually small compared to
microchannel resistance and is therefore neglected.
One challenge for using multiple sensing channels is the electronic
coupling or crosstalk among channels because the electrolyte
electrically connects all channels. When one particle passes
through a microchannel, it generates a resistance change in this
channel. Because all channels are electrically connected, a
resistance change in one channel can cause a current change in
other channels, and in turn induce a voltage change across the
sampling resistors of other microchannels. This voltage change can
be translated into a change in resistance signals of other channels
that do not correspond to passing particles, thereby resulting in
false detections. The placement of measurement electrodes in the
center of microchannels creates an isolation resistor R.sub.ci'
between each pair of microchannels (see FIG. 28) and reduces the
crosstalk.
FIG. 29 shows the result of a typical crosstalk analysis of our
device using PSpice.RTM. (PSpice.RTM. can be obtained from Cadence,
Inc of San Jose, Calif.) assuming R.sub.ci=R.sub.ci'. When the
isolation resistance R.sub.ci' is 10M.OMEGA., the relative
crosstalk (.DELTA.R.sub.2/R.sub.2)/(.DELTA.R.sub.1R.sub.1) in the
adjacent channel (where the crosstalk is maximal) is approximately
0.2% and is considered negligible. As the isolation resistance
increases, the crosstalk is further reduced. When R.sub.ci'=100
M.OMEGA., the cross talk is approximately zero.
The resistance of a microchannel can be estimated by R=.rho.L/A,
where .rho. is the resistivity of the electrolyte, L is the length
of the microchannel, and A is the cross section of the
microchannel. In this work, we use deionized (DI) water, with a
resistivity of about 8.33.times.10.sup.3 .OMEGA.-m, to carry the
microparticles. For the microchannel we used, the estimated
resistance of the DI water filled-microchannel (R.sub.ci') is in
the order of 100 M.OMEGA.. Thus, the crosstalk is negligible. When
the channel size is scaled down to the submicron and nanometer
level, according to the scaling law, R.sub.ci' will be increased
significantly, and thus much less crosstalk is expected. Therefore,
some embodiments having nanoscale channels can operate without
crosstalk even when using a more concentrated electrolyte having a
lower resistivity. This is particularly useful because,
concentrated electrolytes are often necessary for carrying certain
bio-particles.
The microchannels and reservoirs are fabricated on
polydimethylsiloxane (PDMS) using soft lithography, and are bonded
to a glass substrate with sputter gold electrodes. Device layout
(microfluidic channels and electrodes) can be printed onto
transparency films using a high-resolution laser printer. The
transparency films can then be used as masks in contact
photolithography to generate masters with a negative UV photoresist
(MicroChem Corporation XP SU-8 2010, Newton, Mass.) on a glass
slide for the channels.
According to one example, a curing agent and PDMS prepolymer
(SYLGARD 184 Silicone Elastomer Kit, Dow Corning, Midland, Mich.)
are mixed in a 1:10 weight ratio. The prepolymer mixture is
degassed in a desiccator with a vacuum pump for one hour to remove
any air bubbles in the mixture. Then, the prepolymer mixture is
poured onto the master. The master/PDMS stack is cured for three
hours at 80.degree. C. on a hot plate. After curing, the thin PDMS
replicas are cut and peeled off of the masters. Next, contact
photolithography with a positive AZ 4620 photoresist (AZ Electronic
Materials, Somerville, N.J.) is presented on another glass slide to
create the electrode patterns. Cr/Au (50 .ANG./3000 .ANG.) sheet
films are evaporated on the glass slide. A subsequent liftoff
process completes fabrication of the electrodes. The PDMS layer
with developed channels and electrodes-embedded glass slide are
then treated with RF oxygen plasma (Plasma Etcher PE 2000, South
Bay Technology Inc., San Clemente, Calif.) for 25 seconds (50 W,
200 mTorr). This temporarily activates the exposed part of the PDMS
and provides very good adhesion. The PDMS replica and glass side
are then immediately brought into contact, aligned, and bonded
together.
A single measurement channel comprises the half channel resistance
R.sub.ci in series with the sampling resistor R.sub.s and the
supply voltage (see FIG. 27). When a particle passes through the
microchannel it causes a change in the half channel resistance, and
a corresponding change in the voltage across the sampling resistor.
The relative change in resistance of the microchannel in terms of
the measured voltage is given by Equation (1), where V.sub.s.sup.1
is the measured voltage when a particle is present in the
microchannel and V.sub.s is the measured voltage in the absence of
a particle in the microchannel.
For a micro-channel with length L and diameter D (see FIG. 28(b)),
the change in resistance as a particle passes through it is given
by Equation (2), where d is the diameter of the particle, and D and
L are the diameter and length of the micro-channel, respectively.
The equation holds true when (d/D).sup.3<0.1, as is the case in
this embodiment. Thus, the particle diameter can be calculated from
the relative change in resistance according to Equation (3).
An illustration of this embodiment follows. Polymethacrylate
particles with diameters of 40 .mu.m (40 .mu.m.+-.0.8 .mu.m) (Sigma
Aldrich Inc.), and Rocky mountain Juniper (Juniper Scopulorum) tree
pollens (Sigma Aldrich Inc.) are chosen for the following example.
These particles are chosen because they are commercially available
and because polymethacrylate particles have well characterized
properties. The diameters of pollen particles are determined using
high resolution optical microscopy. The Juniper tree pollen is
egg-shaped and the diameter ranges from 17 .mu.m to 23 .mu.m. FIGS.
30(a), and 30(b) show photomicrographs of 40 .mu.m polymethacrylate
particles and Juniper tree pollen, respectively.
The particle solution is forced to flow through microchannels of
the present invention by application of a pressure difference with
a syringe. An applied voltage of V.sub.cc=6V is applied across the
microchannels. Due to the polarization effect of gold electrodes,
such a high source voltage is necessary to ensure that there is
sufficient current/electric field within the electrolyte to record
a noticeable voltage change across the sampling resistors. Voltage
measurements are made across a sampling resistor R.sub.s=100
k.OMEGA.. The voltage trace is recorded for four channels using a
National Instruments NI-6220 data acquisition board, with a
sampling frequency of 50 kHz.
The application of a 6V DC voltage on the electrodes in electrolyte
can, in some cases, cause electrolysis of water and generate gas
bubbles. The gas bubbles can result in false peaks when they pass
through the microchannel. No such bubbles are observed in this
example.
The Juniper pollen particle solution is prepared by diluting 10 mg
of Juniper tree pollen in 10 mL of water. FIG. 31 shows the
relative resistance change of the four channels as a function of
time. Resistance is calculated by converting it from voltage traces
measured across the four sampling resistors using Equation (1).
Each resistive pulse represents one pollen particle passing through
a microchannel. The resistive traces show pulses appearing in
random sequence. A cross correlation analysis is performed between
the signals from two sensing channels at a time. The results are
shown in FIG. 32. These results show that the cross correlation
coefficients |r| are all less than 0.1, indicating there is
negligible correlation among the pulses of different channels.
Thus, the four sensing channels are able to simultaneously detect
and count particles with negligible crosstalk among channels.
It is obvious from FIG. 31 that the resistive pulses caused by
Juniper pollens are all downward, that is, when a pollen particle
passes through the microchannel, the microchannel resistance
decreases. This is possibly because of the high surface charge of
pollen particles. This phenomenon can be explained in terms of the
surface charge of particles, and can be used to distinguish between
particles with different surface charges. Notably, the pulses of
the Juniper pollen vary in height. Although not wishing to be bound
to any one theory, the variation might be attributed to the
particle size variation and the egg-shape of Juniper pollen.
Polymethacrylate particle solution (10% solid) 0.1 mL of 40 .mu.m,
and 10 mg Juniper pollen are mixed in 10 mL DI water and are tested
in a multichannel embodiment. The resulting concentration of 40
.mu.m polymethacrylate particles is 2.49.times.10.sup.4 mL.sup.-1.
Voltage traces across the sampling resistors are recorded. A
typical resistive pulse trace in one channel (channel 3) is shown
in FIG. 33. The resistive trace is converted from the voltage trace
signal. Magnified resistive pulses generated by Juniper pollen and
40 .mu.m polymethacrylate particle are shown in FIG. 34. According
to these results, pollen generates downward resistive pulses, while
polymethacrylate particles generated upward resistive pulses. Thus,
we are able to differentiate and count the two particle species.
The concentration of the 40 .mu.m polymethacrylate particles in the
four channels is calculated from counting the number of upward
peaks during the period of one second. The concentrations are
calculated to be 2.15.times.10.sup.4 mL.sup.-1, 2.11.times.10.sup.4
mL.sup.-1, 2.14.times.10.sup.4 mL.sup.-1, and 2.01.times.10.sup.4
mL.sup.-1 for channels 1, 2, 3 and 4, respectively. These results
are shown in FIG. 35. The measured particle concentration in each
channel is lower than the estimated particle concentration, which
is 2.49.times.10.sup.4 mL.sup.-1. This may be due to some PM
particles depositing onto the substrate during the experiment.
The particle diameters are calculated from resistive pulse data
shown in FIG. 33 using Equations (2) and (3). Using the nominal
sensing microchannel dimension of 50 .mu.m.times.100
.mu.m.times.150 .mu.m, the analysis shows the estimated particle
diameter is 20.1.+-.1.8 .mu.m, 20.4.+-.1.5 .mu.m, 22.0.+-.2.1
.mu.m, and 22.4.+-.1.8 .mu.m for channels 1, 2, 3 and 4,
respectively (see FIG. 35). The large difference between the
calculated and the actual particle diameter (40 .mu.m.+-.0.8 .mu.m)
is mainly because of the polarization effect that takes place on
the gold electrodes. In electrolyte solution and DI water,
electrode polarization causes the DC voltage applied on electrodes
be dropped across the double layers of the two electrodes. Thus,
the voltage drop across the bulk solution is less than the actual
applied voltage, resulting in underestimated particle dimensions
when Equation (3) is used. The electrode polarization can be
reduced by using Ag/AgCl electrodes with large surface areas.
IV. Automated Continuous Aerosol Sampling and Coulter
Counting-Based Bio-Aerosol Detector System Embodiments
Another embodiment includes a rapid, integrated, particle screening
device consistent with the design set forth in FIG. 4. The
embodiment comprises a bio-aerosol sampler for collecting airborne
pollen and combining it with a liquid, such as DI water. The
embodiment also includes a multichannel Coulter counter for rapid
detection and counting of particles. According to one version of
this embodiment, an air sampling system can be included. Some
embodiments can also include continuous automated aerosol
monitoring. One very specific embodiment includes the Biosampler
air sampling system from SKC Inc. Some embodiments are self
cleaning and can be integrated with a Coulter counter. According to
some embodiments, and in keeping with FIG. 4, a sampling bottle can
be connected to a reservoir of DI water, and a flow of liquid from
the bottle can be directed to a multichannel Coulter counter. Some
embodiments are also capable of recycling the DI water and/or
self-cleaning the DI water, thereby extending the potential
duration of unattended operation. According to some embodiments an
air sample can be pumped through a particle separation and/or
collection pipe.
According to one embodiment anthrax spores can be detected in mail
during processing/sorting of the mail. Some embodiments may offer a
cheaper and/or more rapid detector of anthrax in mail room
applications. FIG. 36 shows a drawing of one example of this
embodiment. The envelopes on the mail sorting grill are subjected
to a generally perpendicular air flow from, for instance, a blower.
The particles on the envelopes, including medium size anthrax
spores (a few microns), large dust particles (a few tens of microns
or larger) and small dust particles (sub microns or less) are swept
off the envelops and carried by the air flow to one or more
collection devices.
En route to the collection device, the air flow is subjected to a
sharp turn through a flow channel. The larger and/or heavier
particles, such as dust particles, are separated and exit at an
elbow or bend in the channel, while the medium sized and small
sized particles continue toward the collection device. FIGS. 37(a)
and 37(b) show fluid mechanics simulation results of particle
behavior in such embodiments where the inlet velocity is about 10
m/s. Specifically, FIG. 37(a) shows that particles having diameters
of about 1 mm are separated at the elbow. The dots that depart from
the central line in FIG. 37(a) represent the separating particles.
Heavier/larger particles tend to separate, while smaller/lighter
particles tend to be guided by the air flow. FIG. 37(b) shows the
average path of particles having diameters of about 1 micron.
Next, the air flow is directed to, and impinges, a moving
collection belt. Mid-sized particles, including anthrax, adhere to
the surface while smaller particles remain in the air flow, which
turns parallel to the collection belt. The collection mechanism is
similar to known particle sorting methods, such as virtual
impactors. In one embodiment, the movable belt transports captured
particles to an ethylene glycol tank. The tank both collects the
particles, and provides continuous wetting of the collection belt
thereby enabling particle capture. Unlike water based electrolyte
solutions where water evaporates faster, using ethylene glycol as
the electrolyte diminishes the need to constantly compensate
evaporative losses.
After the particles are collected, the solution is transported to
the first stage multichannel Coulter-type sensor. Because of the
relatively large volume of analyte, a multichannel Coulter counter
is used thereby enabling rapid analysis of analyte solutions.
Anthrax spores are heavily charged bioparticles. Therefore, the
first stage Coulter counter is able to distinguish them from dust
particles without the need to add costly antibodies.
It is desirable that anthrax detectors minimize false positives.
Therefore, some embodiments can include a second stage Coulter
sensor to further diminish the chances of false positives. If the
first stage sensor detects suspicious particles, the second stage
verifies whether the particles are anthrax. In one embodiment, the
second stage includes a valve that can be used to direct
electrolyte solutions containing suspicious particles to a second
stage sensor. According to some embodiments, bioselectivity can be
enhanced by using monoclonal antibodies that bind to the anthrax
spore specifically. Binding to a monoclonal antibody causes a
change in the size and surface charge of anthrax spores, which can
be identified by the Coulter counter detector.
Some embodiments are capable of detecting a single anthrax spore in
a liquid analyte. Some embodiments may also include multiplexing
technology that significantly improves the signal-to-noise ratio of
the analytical signal.
Some embodiments are suitable for mass production and can comprise
a low-cost and/or portable device. Some embodiments may be suitable
for rapid on-site aerosol particle analysis. Some embodiments may
require no complex setup, and/or no enrichment of the analyte prior
to contacting the detector, or other preparation of the
analyte.
Some embodiments can comprise high throughput bio-aerosol
instrumentation. Some embodiments can comprise a large number of
channels in one chip and thus allows real-time detection of anthrax
spores.
According to some embodiments, the detector is capable of counting
each spore, bacterium, and/or virus in the collection electrolyte
solution. Thus, sensitivity can be below 100 CFU/liter. Some
embodiments can deduce size, shape, and/or surface charges based on
current measurements with greater than 99% confidence. According to
some multi-channel embodiments detection efficiency and/or response
time is greatly improved. For example, response time can be less
than one minute. Furthermore, some embodiments comprise a
substantial dynamic range, enabling detection of a wide variety of
sizes of particles. According to some embodiments, integrated
micromachined devices can allow small size (e.g. <1 cu. ft) and
low unit cost (e.g. about $1000/unit). Some embodiments can enable
high reliability, low maintenance and/or operating cost (e.g. less
than about $1000/year).
Some embodiments are capable of the same or better selectivity as
compared to immunoassays. However, unlike most common immunoassay
methods, which require antibodies to be labeled (e.g. using
fluorescence, radioactivity, or enzyme activity), some embodiments
are capable of operating without such labels. Thus, the device is
relatively inexpensive to use. According to some embodiments,
operating costs can be limited by using a two-stage sensor system,
where the second stage (i.e. the antibody stage) is active only
when the first stage sensor sends an alarm signal indicating a
suspicious signal has been detected.
V. Multiplexed Multichannel Coulter Counter Embodiments
Some embodiments comprise piezoelectric bimorph microactuators.
Such actuators can be bonded, for example, to the outside of each
microchannel to dynamically control the flow, thereby modulating
the electrical impedance of the microchannel. In some embodiments,
bimorph actuators can be a cantilever made of two different
materials bonded together in such a way that when one layer deforms
in response to an applied voltage, the cantilever bends. In some
embodiments, the bimorph actuators also allow the effective size of
the microchannel to be changed, for instance, to match the size of
tested particle, thereby improving the range of particle sizes that
the Coulter counter can detect.
One embodiment comprises a four-channel Coulter counter, as shown
in FIG. 38. The four-channel Coulter counter comprises two chips: a
microfluidic chip and actuator array chip. In the microfluidic
chip, reservoirs and microfluidic channels can be microfabricated
by depositing polydimethylsiloxane (PDMS) to form desired
geometries. Two electrodes can be located in the two major
reservoirs for applying a constant voltage across the channels
and/or for taking measurements. Electrolyte solutions containing
particles are forced to move from the inlet reservoir to the outlet
reservoir through the four sensing channels. In order to
dynamically modulate the microchannels, the bottom channel wall can
comprise a thin (e.g. about 50 .mu.m) PDMS membrane. PDMS is a soft
material (Young's Modulus of cured 10:1 Sylgard 184 from Dow
Corning is about 2.5 MPa) and it can be easily deformed. PDMS
microchannels can be modulated with mechanical pins actuated by
piezoelectric bimorph.
In order to modulate the channel signals, a microactuator array can
be bonded to the microchannels and used to mechanically modulate
the sizes of the microchannels. FIG. 39 shows a schematic view of
an actuator cell, which comprises a single crystal PMN-PT/Silicon
bimorph actuator forming the active layer, and a pin structure as
the linkage to the microchannel. In some embodiments the bimorph
actuator can comprise a plurality of layers including two, three or
four layers. In embodiments having four layers, the PMN-PT single
crystal can be sandwiched between two electrodes (not shown), atop
which is single crystal silicon (e.g. with a thin oxidation
layer).
According to some embodiments, the pins are fabricated onto the
bimorph actuator at the cantilever ends. The PMN-PT layer can be
active in the sense that when a voltage is applied it deforms
axially. Since the silicon and PMN-PT layers are sandwiched
together, the free axial deformation of the top layer (e.g. Si) is
constrained by the substrate layer (e.g. PMN-PT) and, as a result,
the composite beam bends. The resulting deformable shape of the
beam is the one shown in FIG. 39(b). The bending is converted to a
normal displacement on the pin, which in turn causes deformation of
the microchannel wall. In general, the fast response of
piezoelectric actuators enables the microchannels to be modulated
on and off with very high frequency. In turn this may enable the
multiplexing of measurement signals as a particle traverses a
channel.
According to some embodiments, the microactuator array is
independent of the microfluidics and thus does not disturb the
electrical measurements. By accurate control of the voltage applied
to individual actuator cells, the microchannel membrane can also be
deformed to decrease the size of the microchannel rather than
closing it completely, so as to allow for the detection of smaller
particles (e.g. compare FIGS. 39(b) and 39(c)).
Unlike the parallel multichannel devices that measure individual
current traces for each channel, some embodiments are capable of
measuring only the current through the entire system. Thus,
according to these embodiments, the measured current is the sum of
the response of multiple channels. For example, when a
microactuator closes a channel, it blocks the ion current from
flowing through the channel. If the microchannels are opened and
closed selectively (e.g. in a pattern specified by a Hardmard
transform), it is possible to recover the individual channel
currents from the measured total system current. According to this
embodiment, the ability to close and open channels at a high enough
frequency enables the application of the Hardmard transform. Thus,
high-frequency switching of the actuator can be used to take
several measurements in the time it takes a particle to traverse a
channel. The number of channels can be multiplexed in this way
depends on the response time of the actuators, the speed of the
data acquisition system, and the travel velocity of the particles
in the microchannels. Thus, faster actuators enable higher numbers
of channels.
According to one embodiment a biomorph actuator can comprise one or
more approximately 500 .mu.m long cantilevers. The cantilevers can
comprise a single crystal piezoelectric active layer, and a silicon
support layer. According to one embodiment, the boundary conditions
can include a cantilever beam clamped at one end and free at the
other. The electrode layers can be thin, are therefore not
necessarily taken into account. FIG. 40 shows a set of plots of
maximum cantilever deflection versus applied voltage for the five
different cantilever designs. The calculated stiffness and natural
frequency of the bimorph structures are also shown in the
figure.
Analyses such as that which is shown in FIG. 40 can be used for
estimating actuator performances. FIG. 40 indicates that a
PMN-PT/Si bimorph cantilever can be capable of more than a 10 .mu.m
tip deflection, which is enough to turn a microchannel off. For
example, for a 500 .mu.m cantilever with 20 .mu.m thick PMN-PT and
a 20 .mu.m Si substrate, 14 .mu.m deflection is attainable at 50 V,
while the natural frequency is 66.7 KHz. For a low voltage design
using 10 .mu.m thick PMN-PT and a 5 .mu.m Si substrate, an 18 .mu.m
deflection is predicted at 10 V. These results indicate that it is
possible for a bimorph actuator to fully mechanically close a
channel with a relatively low voltage.
Some embodiments are capable of preventing channel wall rupture by
the actuator. According to one embodiment a PDMS channel wall
rupture can be avoided by including a corrugated membrane.
According to some embodiments such membranes can be fabricated by
micromolding.
As would be readily appreciated by one of ordinary skill in the
art, the present invention is not limited to PMN-PT actuators, but
rather can include any of a wide variety of macro and/or micro
actuators that have sufficiently fast responses.
According to some embodiments the foregoing actuators can be used
for a wide variety of applications including, without limitation,
drug delivery and microfluidic control.
VI. Polymer Brush Conductance Modulation Embodiment
According to another embodiment the conductance within a
microchannel is modulated using polymer brushes grown within
microfluidic channels. In this regard, the stimuli-responsive
behavior of grafted polyelectrolyte brushes in aqueous solution
under the influence of an electric field is employed to modulate
conductance. A gated function is realized as the polymer brushes
transition from an erect state to a collapsed state in response to
an external electric field. Based on the same basic design shown in
FIG. 38, the polymer brushes are used to modulate individual
channels within the device.
FIG. 41 shows the basic design for the polymer brushes. FIG. 41(a)
shows the polymer brushes in a collapsed state, and FIG. 41(b)
shows the polymer brushes erect or extended in response to a
negative electrical field. FIG. 42 provides a simulated view of the
three states of freely jointed bead-chain polymer brushes:
equilibrium, collapsed and erect. The transition between states
requires a critical intensity of the electric field to induce the
response and trigger the transition. Under a positive electric
field, the polymer chains relax and collapse, forming irregular
molecular groups on the channel wall, as seen in FIG. 42(b).
Conversely, a negative electric field causes the chains to stretch,
as shown in FIG. 42(c). By controlling the electric field
direction, or polarity, the transitioning of the polymer chains
from erect to collapsed is also repeatably controlled. FIG. 43
provides a graph of the foregoing, plotting average height change
of the polymer brush, normalized by Lennard-Jones (L-J) diameter
(.sigma.=4 .ANG.) against electric field change as a function of
time in .tau.(.tau.= {square root over (m/.epsilon.)} where m is
the mass of one chain segment and .epsilon. is the L-J unit of
energy.
Based on the foregoing, the morphology of gated polyelectrolyte
brushes can be perturbed using electric signals, providing a means
to realize biomimetic fluidic/ion channels for a number of varying
applications. One such application involves a voltage-gated ion
channel suitable for use in coordination with the nervous system.
In this instance a number of transmembrane protein helices act
coordinatively as a gate and are the primary determinant of the
high ion selectivity and precise control of the conductivity. Based
on this function, which mimics that of the fluidic/ion channels in
biological systems, a new class of advanced gating systems for drug
delivery, bioimplants, data storage, smart valves, nano electric
devices, and other similar biological applications may be
generated.
Electroactive polymer brushes are typically anchored to Au electric
contacts by thiol or disulfide based surface anchors. However, such
linkages are susceptible to physical and electrochemical
degradations, usually within 10 s cycles of cyclovoltammetric
scanning. Oxidation of the thiol to sulfonates is attributed to the
rapid loss of surface functional groups. To increase the stability
of the polymer brushes suitable for repetitive electrochemical
modulation of the mechanical properties of the brushes, a special
surface anchor based on the 2,4,7-trithia-tricylo[3.3.1.1.sup.3.7]
decane moiety was prepared.
To synthesize the polymer brushes in accord with the disclosure
provided, surface polymerizations were performed using the
following reaction scheme based on atom transfer radical
polymerization (ATRP). The formation of the desired polymer brush
layer on the electrode surface using this polymerization scheme was
confirmed by surface IR spectroscopy, as set forth in FIG. 44. In
this Figure, a PM-FT-IRRAS analysis of the hydrocarbon region of
the polymer brush growing on a Au surface is provided. The larger
trace is the polymer brush after 9 hours reaction and the smaller
trace is the polymer brush after 3 hours.
##STR00001## VII. Amplitude Modulation Embodiments
In another embodiment a multiplexed signal representing multiple
channels is acquired and demodulated to recover individual channel
signals. In this embodiment, a single pair of sensing electrodes
and a single DAQ (Data Acquisition Board/System) channel are needed
to recover multiple signals. The multiplexed multi-channel sensor
design is similar to that presented hereinabove for example in FIG.
38, and more specifically FIG. 45, but a different measurement
scheme is used. Specifically, major electrodes are positioned, one
on each side of the sensing channels, and used to make a combined
measurement with the multiplexed response of all channels.
In another embodiment, next generation integrated lab-on-a-chip
devices for detection and quantification of important biological
targets, including bacteria, viruses and DNA sequencing, urgently
needed in public health monitoring and biomedical research, are
developed. In order to provide such devices for field application,
high density parallel sensing arrays that are capable of processing
large volumes of sample in a reasonable time are needed. As has
been established hereinabove, coulter counters are known tools for
sizing and counting cells in colloidal particles. Such devices,
however, are limited in the fact that they have relatively low
throughput; such devices are fundamentally serial devices that scan
particles individually as they pass through a microchannel, making
real-time analysis of bulk fluid samples difficult.
In response to this need, in one embodiment, devices with parallel
micro-channels, each channel being equipped with individual
detection electronics, are provided. In such devices, each channel
can essentially be considered an individual instrument. Multiplexed
detection is necessary in devices where the number of channels is
large and individual detection electronics are impractical.
Therefore, by modulating signals from the individual channels
differently, a multiplexed signal representing a number of channels
can be acquired, and then that signal can be demodulated to recover
the individual channel signals.
FIG. 45 shows the design concept of a multi-channel
micro-fabricated coulter counter. The device consists of two
reservoirs to load analytes, four parallel microfluidic channels,
each being 400 .mu.m.times.50 .mu.m.times.40 .mu.m, for counting
particles, a pair, or one set, of main detection electrodes, one on
each side of the microchannels to apply a constant DC bias
(V.sub.cc), and a central electrode in each microfluidic channel
exposed only at the center to apply AC modulation signals. In this
design, each central electrode divides the micro-channel in to two
half microchannels. Devices of this design may be created by
lithographic or micro-machine techniques in accord with the
disclosure presented hereinabove. For example, a PDMS substrate can
be micro-machined using soft lithography to create a device in
keeping with the techniques disclosed herein or other suitable
techniques.
FIG. 46 provides a simplified equivalent circuit and measurement
scheme for the four-channel device discussed above. Each half of a
microchannel is modeled as a resistance (R.sub.ci & R.sub.ci').
The current through each microchannel flows into a unity gain
inverting summing amplifier through the center electrodes. The
combined response, V.sub.s, is monitored at the output across a
sampling resistor R.sub.s. The signal in each channel is modulated
by an AC sine-wave of known and unique frequency (amplitude
modulation). The combined response is demodulated to obtain the
signals from each individual channel.
30 .mu.m polystyrene particles suspended in 0.1M NaCl was used to
test the device. Each of the central electrodes was connected to a
sinusoidal AC voltage source to apply a modulation signal,
specifically, 300 mV peak-to-peak, using a different frequency for
each channel. The particle solution was injected through the inlet
reservoir using a pressure-driven flow. The combined response,
V.sub.s, was measured at the output of the amplifier across a
sampling resistor, R.sub.s. The response was recorded at a sampling
rate of 1 MHz using a NI PCI-6133 data acquisition board (DAQ) and
a LabView interface.
FIG. 47 provides a block diagram of the signal processing steps
used to modulate the individual channel signals to produce a
combined signal, and then demodulate the combined signal to recover
the individual channel signals. The combined response (V.sub.s)
measured across the sampling resistor (R.sub.s) was demodulated by
filtering out all but the lower side band (LSB) for each channel.
The filtering was accomplished using a rectangular window band pass
filter based on a Fast Fourier Transform (FFT). To demodulate the
signal for a channel modulated at frequency f.sub.c, signal content
in bandwidth varying from f.sub.c-5 kHz2f.sub.c to 0.5 kHz was
retained. Envelope detection was used on the resulting signal to
remove the AC sine wave carrier and thus recover the signal for
each individual channel. The filtering and envelope detecting were
implemented using custom code written in MATLAB.
Piezoelectric micro-actuators can be used for dynamic modulation of
micro-fluidic channels. To dynamically modulate multiple
microchannels typically requires power supplies of high modulation
frequency (typically 50 kHz) and high actuation voltage (>100V).
However, the application of these power supplies to microfluidic
devices is impractical and costly, which contradicts the goal of
development of miniature, low cost bio-instruments.
In contrast, electrical-field-effect polymer brushes grown as a
nanofilm on the surface of microchannels can switch between the
extension and collapse status with a low applied driving voltage.
The fast extension/collapse switching allows dynamic modulation of
the conductance/impedance of electrolyte-filled microchannels. Thus
the signals from each of the channels can be demodulated from the
overall signal to produce the total counting/detecting signal
without much increase in device complexity and power
consumption.
Signal recovery using amplitude modulation was demonstrated using
measurements first made for a single channel device using both DC
mode and amplitude modulation mode. A DC voltage (V.sub.cc=1 V) was
applied across the microchannel and an encoding AC signal (300
mVp-p, f.sub.c=22 kHz) was applied through the central electrode.
The combined response was recorded. The DC response was obtained by
retaining the low frequency response (0-5 kHz) by filtering out the
high frequency (<5 kHz) components from the measured response
and the demodulation step, whereas the amplitude modulated response
is obtained by retaining the LSB components with the bandwidth
varying from 17 kHz to 21.5 kHz for f.sub.c=22 kHz. The pulses of
the demodulated response occur at the same times as those of the DC
response, as set forth in FIG. 48. This FIG. 48 provides test
results with the DC offset removed for 30-.mu.m polystyrene
particles passed through a single channel counter, V.sub.cc=1V. In
this figure, (a) shows the DC response, while (b) provides the
demodulated response. Each pulse shown represents one particle
passing through the microchannel. This test demonstrates the
feasibility of particle detection using the amplitude modulation
method.
Following on the foregoing, a multiplexed measurement was done on
two channels of the four-channel device. The two channels were
modulated at frequencies f.sub.c1=20 kHz and f.sub.c2=65 kHz.
Recovered signals for individual channels are shown in FIG. 49.
FIG. 49 provides measurements for the two channels of the
four-channel device wherein V.sub.cc=1V, (a) shows channel 1 having
encoded frequency f.sub.c1=20 kHz and (b) shows channel 4 having
encoded frequency f.sub.c2=65 kHz. Each pulse represents one
particle passing through the identified microchannel.
FIG. 50 illustrates details of the signal processing used to
recover the signal for the first channel from the combined signal
that is the output of the sensor. FIG. 50(a) is a selected section
of the filtered signal for the LSB for channel 1 from FIG. 49
showing a variation in voltage as a particle passes through the
channel with superimposed output of the envelope detector detecting
the variation. FIG. 50(b) shows output of the envelope detection
showing a voltage pulse for channel 1. FIG. 50(a) then shows the
signal after applying the LSB filter to isolate the part of the
combined signal that relates to the carrier frequency for the first
channel. The circle data points on the plat indicate the results of
envelope detection, which is used to remove the carrier signal. The
output of the envelope detector, shown again in 50(b) is the
recovered signal for the first channel.
It is noted that a change in the current of one channel has the
potential to effect the response of a neighboring channel, thereby
resulting in false detection of particles in the neighboring
channel. Cross correlation analysis between the two channel
signals, shown in FIG. 51, however, shows that the cross
correlation coefficient is |r|<0.01 indicating cross talk
between the two channels is negligible.
Using the synthesis method described above, a surface anchor that
can attach the polymer brushes to the electrode stably was
synthesized. Cyclovoltammetry (CV) measurement results are
presented in FIG. 52:
Specifically, FIG. 52 provides CV measurements obtained in a three
electrode electrochemical cell employing the Gamry.RTM.
Electrochemistry Workstation under the following conditions: 1.0 mM
K.sub.3[Fe(CN.sub.6)]/K.sub.4[Fe(CN.sub.6)], as the reversible
redox pair, 0.1 M KCl as the supporting electrolyte, a Au electrode
as the working electrode, a Pt wire as the counter electrode, and a
Ag/AgCl electrode as the reference. (a) bare gold electrode; (b)
initiator self-assembled monolayer (SAM) modified gold
electrode.
In FIG. 52, after the gold surface was modified with the anchor and
the polymer brush initiator, no redox peak of probe molecules and
only very small response current was present (curve b), indicating
good surface integrity of the SAM layer.
FIG. 53 CV measurements obtained under the following conditions;
1.0 mM K.sub.3[Fe(CN.sub.6)]/K.sub.4[Fe(CN.sub.6)] as the
reversible redox pair, 0.1 M KCl as the supporting electrolyte, a
Au electrode as the working electrode, a Pt wire as the counter
electrode, and a Ag/AgCl electrode as the reference. (a) gold
electrode modified with initiator SAM; (b), (c), (d), (e), (f) SAM
modified gold electrode after being treated with 0.1 M acetonitrile
for 15, 30, 45, 60, 90 minutes; (g) bare gold electrode.
Next, acetonitrile, a well-known competitive ligand which typically
damages the thiol based surface SAM, was used to treat the
initiator SAM attached on the working electrode. The CV
measurements were taken every 15 minutes. FIG. 53 shows that after
the SAM was treated with acetonitrile for up to 90 minutes, there
was no redox peaks of probe molecules, suggesting that the
initiator SAM has good surface integrity and stability.
FIG. 54 provides the CV measurements after the SAM layer was
treated for 15 minutes in acetonitrile at different concentrations
from low to high. It is obvious that no redox peak appeared after
the SAM was treated with high concentration (0.5M and 11.0M)
acetonitrile. This further indicates the surface integrity and
stability of the initiator SAM as synthesized.
Finally, surface polymerizations were performed with several new
reaction conditions. Surface IR spectroscopy (PM-FTIRRAS) analysis
was conducted. The initial experimental results, shown in FIG. 44
indicate that the desired polymer brush nanofilm was formed on the
gold electrode surface successfully.
The foregoing examples are considered only illustrative of the
principles of the invention rather than an exclusive list of
embodiments. Further, since numerous modifications and changes will
readily occur to those of ordinary skill in the art, the invention
is not intended to be limited to the exact construction and
operation shown and described, and accordingly, all suitable
modifications and equivalents are within the scope of the present
invention.
* * * * *